laboratory techniques tissueextractionof · stat frozen sections (8 pm) or paraffin wax...

J Clin Pathol 1990;43:499-504 499

TECHNIQUES

Tissue extraction ofDNA and RNA and analysisby the polymerase chain reaction

D P Jackson, F A Lewis, G R Taylor, AW Boylston, P Quirke

AbstractSeveral DNA extraction techniques were

quantitatively and qualitatively com-pared using both fresh and paraffin waxembedded tissue and their suitabilityinvestigated for providing DNA and RNAfor the polymerase chain reaction (PCR).A one hour incubation with proteinase

K was the most efficient DNA extractionprocedure for fresh tissue. For paraffinwax embedded tissue a five day incuba-tion with proteinase K was required toproduce good yields of DNA. Incubationwith sodium dodecyl sulphate producedvery poor yields, while boiling produced20% as much DNA as long enzyme diges-tion. DNA extracted by these methodswas suitable for the PCR amplification ofa single copy gene. Proteinase K digestionalso produced considerable amounts ofRNA which has previously been shown tobe suitable for PCR analysis. A delaybefore fixation had no effect on theamount ofDNA obtained while fixation inCarnoy's reagent results in a much betterpreservation ofDNA than formalin fixa-tion, allowing greater yields to be extrac-ted.

Department ofPathology, Universityof Leeds, Leeds, LS29JTD P JacksonF A LewisA.W BoylstonP QuirkeYorkshire RegionalDNA Laboratory,Leeds GeneralInfirmary, LeedsG R TaylorCorrespondence to:Dr P QuirkeAccepted for publication16 January 1990

The molecular biological analysis of diseaseprocesses requires the extraction ofDNA andRNA from a variety of tissue preparations,including fresh and paraffin wax embeddedmaterial. The extraction of nucleic acids fromroutinely fixed and processed tissue is par-

ticularly important as it permits the use ofarchival material for the retrospective inves-tigation of disease.Techniques for the extraction ofDNA from

fresh tissue and cytological preparations havebeen amply described'2 and since adapted foruse with paraffin wax embedded tissue. Thedisadvantage is that these methods requirelarge amounts of tissue to produce sufficientDNA suitable for analysis by standardmolecular biological procedures."4RNA is highly labile and ribonucleases

(RNAses) are endogenous in tissues andubiquitous in the environment and thereforerequire more complex extraction methods.5These have not been described for use withfixed, processed material.

Fixation and embedding of tissue is a

procedure that has a profound effect on itsmolecular arrangement. It is therefore neces-

sary to investigate what effect this processinghas on the ease with which DNA can beextracted from tissue.

The polymerase chain reaction (PCR) is arapid technique for the enzymatic, in vitroamplification of DNA,6 or RNA, by the addi-tion of a reverse transcription step.78 Thetechnique is very sensitive and can toleratesmall quantities of poor quality template DNAor RNA. This has important implications as itmay permit the use of very rapid simple DNAand RNA extraction techniques to providematerial for PCR analysis.We performed a quantitative and qualitative

study, comparing several reported DNAextraction procedures for use with both freshand paraffin wax embedded tissue, and inves-tigated their suitability to provide DNA andRNA for PCR. We also investigated differentaspects of tissue fixation and processing todetermine their effects on DNA extraction.

MethodsFresh tonsil, obtained immediately afterremoval, was used as a model tissue for DNAextraction experiments. The tissue was cut into100 mg blocks for comparative experiments, or

blocks suitable for sectioning, and either frozenin liquid nitrogen or fixed and then processedon a Shandon 2LE automatic tissue processorbefore being embedded in paraffin wax. Cryo-stat frozen sections (8 pm) or paraffin wax

embedded sections (4 gm) were cut on to glassslides. Tissue was fixed in 4/O" formaldehyde(tap water formalin), neutral buffered for-malin, formol sublimate, paraformaldehyde orBouin's reagent.9 Archival splenic tissue was

obtained from the archives of the Departmentof Pathology, University of Leeds.Owing to the impossibility of absolute con-

trol over variables such as cellularity, area, andthickness within fresh and paraffin wax embed-ded tissues, together with the amount of tissuerequired for all of the different studies perfor-med, different blocks of tissue were used foreach experiment. Tonsilar or splenic tissue wasused throughout, the blocks of which were cuton the same microtome where possible. Serialsections were always used and taken from thesame blocks of tissue for comparison. Whenweight of tissue was used blocks were takenfrom the same tonsil. Each extractionprocedure was routinely performed on 10tissue sections.

DNA ASSAYSExtracted DNA samples were assayed quan-

titatively using the TKO-100 dedicated mini-fluorometer (Hoefer Scientific Instruments,San Francisco, USA) which measures the

LABORATORY

499J Clin Pathol 1990;43:499-504

on March 9, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.43.6.499 on 1 June 1990. Dow

nloaded from

http://jcp.bmj.com/

Jackson, Lewis, Taylor, Boylston, Quirke

fluorescence of Hoescht 33258 (Polysciences,Warrington, Philadelphia) in the presence ofDNA. Human placental DNA (ClontechLaboratories, Palo Alto, California, USA) at100 ng/pl was used to calibrate the mini-fluorometer.The quality of the DNA was determined by

running samples on a 0-70/o agarose electro-phoresis gel (Seakem GTG, FMC bio-products, Uppsala, Sweden), stained withethidium bromide (Sigma, Uppsala, Sweden),in TRIS-acetate EDTA (TAE) buffer in aGNA-100 electrophoresis cell (Pharmacia,Uppsala, Sweden) at 50V for 10 minutes andIOOV for one hour. A A' DNA Hind III digest(BRL) provided a standard, with a ladder offragments: 23130 base pairs, 9416 base pairs,6562 base pairs, 4368 base pairs, 2322 basepairs, 2027 base pairs, 564 base pairs and 125base pairs.

DNA EXTRACTIONThree methods were investigated: proteinaseK incubation modified from the method ofStrauss' and sodium dodecyl sulphate (SDS)incubation, and boiling tissue modified fromthe method of Lench et al.'0

Proteinase K incubationFrozen fresh tissue blocks were crushed to afine powder using a mortar and pestle and thensuspended in 1 2 ml/100 mg tissue of digestionbuffer (100 mM NaCl, 10 mM TRIS-Cl,25 mM EDTA, 0-50o SDS, pH8-4), contain-ing 01 mg/ml proteinase K (Sigma). Paraffinwax embedded sections were dewaxed in warmxylene and rehydrated by passage throughgraded alcohols. The sections were thenscraped into microfuge tubes, suspended indigestion buffer (about 200 pl/ 10 sections), andincubated at 37°C for a variable period. Nucleicacid was purified by an organic extraction step.An equal volume of phenol:chloroform iso-amyl alcohol (25:24:1) was added to theproteinase K digests, mixed thoroughly, andcentrifuged to separate the aqueous andorganic layers. The upper aqueous layer con-taining the nucleic acid was collected and theextraction repeated. A final extraction wascarried out using an equal volumeof chloroform:iso-amyl alcohol (24:1). Thenucleic acid was then precipitated with twovolumes of cold ethanol (- 20°C) and one tenthvolume 3M sodium acetate (pH 5 2) at - 20°Cfor one hour, spun down (13000 rpm, for 10minutes, dried, and redissolved in TRIS-EDTA buffer for between three and five days.

SDS incubationThe tissue was prepared as above, butsuspended in 10% SDS (about 200 Ml//10 sec-tions) and incubated at 50°C for between oneand four hours. The nucleic acid was thenpurified by phenol/chloroform extraction andethanol precipitation.

BoilingThe tissue was suspended in distilled water(about 500 pl/10 sections) and boiled for avariable time in the Perkin-Elmer Cetus (PEC,Norwalk, Connecticut, USA) Thermal cycler.

The resulting solution was used directly foranalysis.

RNASE/DNASE DIGESTIONExtraction nucleic acid samples were digestedwith deoxyribonuclease (DNAse) and ribo-nuclease (RNAse). RNAse A (Sigma) wasdissolved in SSC (0 3 M sodium citrate,30 mM sodium citrate, pH 7 0) to a concentra-tion of 10 mg/ml. This was boiled for 10minutes to destroy any DNAse activity andthen diluted to 200 ,g/ml with SSC. RNAse T,(Sigma) was added to a final concentration of200 ,/ml. DNAse I (Sigma) was dissolved in0-01 M TRIS-Cl containing 0.01 M MgCl2and then diluted down to 200 Mg/ml. A 10 Mlaliquot of each DNA sample was incubatedwith an equal volume of each enzyme prepara-tion for one hour at 37°C and the undigestednucleic acid purified by phenol/chloroformextraction and ethanol precipitation. The sam-ples were analysed by gel electrophoresis (asabove).

POLYMERASE CHAIN REACTIONTwo primers for the human factor VIII genewere used." Fifty picomoles of each wereincubated with 1 pl of extracted DNA samplein 25 Ml of 1 x PCR buffer (50 mM KCl,10 mM TRIS-Cl, 1 5 mM MgCl2, 0.010ogelatin, pH8 3) for five minutes at 94°C in thePEC Thermal cycler to denature fully thetemplate DNA and to inactive any proteasescontaminating the DNA before Taqpolymerase was added. To this was added 25 plof 1 x PCR buffer, containing 2 5 U of"Ampli Taq" (Cetus) and 0 4 mM each ofdATP, dCTP, dGTP and dTTP. The reactionmixture was then incubated for 40 cycles of 48seconds at 94°C, 48 seconds at 50°C, and oneminute at 72'C, followed by 10 minutes at 72°Cto ensure that all the product was fully exten-ded. Each PCR experiment included a negativecontrol of distilled H20 in place ofDNA. ThePCR products were analysed by running on a20,, agarose electrophoresis gel, stained withethidium bromide in a DNA subcell (BioRad)at 150 V for one and a half hours. A 123basepair ladder (BRL) was used as a standard.

ResultsFRESH TISSUEDNAReproducibility: The accuracy of repeat extrac-tions of DNA was assessed both on blocks andsections of fresh tissue. Twelve tissue blocksamples and 12 sets of 10 frozen sections wereextracted by proteinase K incubation. Themean yield and standard deviation for blockswas 6 88 (1 60) pg/mg and for tissue sections12-51 (2 73) pg/mg, confirming the validity ofusing duplicate or triplicate samples.Proteinase K: Most protocols for the extractionof DNA from fresh tissue or cultured cellsrequire tissue to be incubated with proteinaseK for 12-24 hours.' An incubation time of 18hours for the proteinaseK extraction techniquewas a very efficient procedure, capable ofextracting high molecular weight DNA (morethan 20 kilobases) from as little as one frozensection of fresh tonsil. Decreasing the time to

500



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

501Tissue extraction ofDNA and RNA and analysis by PLR

Figure 1 0O7% agaroseelectrophoresis gel ofextracted DNA samples.(a) DNA fromfreshfrozen tonsil sections,extracted by differentmethods. (1) StandardADNA Hind III digest;(2) proteinase Kincubation (18 hours);(3) proteinase Kincubation (one hour);(4) SDS incubation (onehour); (5) boiling (15minutes).(b) DNA from paraffinwax embedded tonsilsections extracted bydifferent methods: (1)proteinase K incubation(five days);(2) SDSincubation (one hour);(3) boiling (15 minutes);(4) standard ADNA HindIII digest.

1 2

1 2 3 4 5 6 7 8 9 10 11

3 4 1 2 3 4

just one hour did not significantly reduce theamount ofDNA recovered, nor did lengthen-ing the incubation time to five days. For thismethod, a proteinase K concentration of0 1 mg/ml was most efficient.SDS: The extraction ofDNA from a cell pelletby a short (one to two hours) incubation with100 SDS has been described.'0 When applied tofresh frozen sections, this method producedrelatively poorer yields than the proteinase Kincubation method. These were only margin-ally improved by increasing the incubation tofour hours.Boiling: Boiling the cell pellet has also beendescribed as an extraction method for freshcytological preparations.'0 Using this methodon frozen tissue, the optimal conditions wereboiling sections for just 10-15 minutes as

increasing the boiling time reduced the yieldsof extracted DNA.Comparison offresh tissue DNA extraction tech-niques: A separate experiment was performed topermit direct quantitative and qualitative com-parison of these different procedures for theextraction of DNA from fresh frozen tissuesections (table 1, fig la). This showed that a onehour incubation with proteinase K producedhigh yields of high molecular weight DNAcompared with the poorer yields of lowmolecular weight DNA (less than 9 kilo-daltons) produced by SDS incubation and thesmall amounts offragmentedDNA from boiledtissue.

RNAThe low molecular weight material produced

Figure 2 0 7% agarose electrophoresis gel ofDNAsamples extractedfrom waxed or dewaxed sections ofparaffin wax embedded tonsil by different lengthincubations with proteinase K.(1) Standard: ADNA Hind III digest.Waxed: (2) one hour; (3) three hours; (4) 24 hours;(5) 48 hours; (6) five days. Dewaxed: (7) one hour;(8) three hours;(9) 24 hours; (10) 48 hours; (11) fivedays.

by these extraction procedures from fresh tis-sue may contain contaminating RNA as well as

fragmented DNA. Digestion with RNAseshowed that both the one hour and 18 hourproteinase K incubations produced substantialamounts of RNA; SDS incubation alsoproduced some RNA; while the small amountof low molecular weight material produced byboiling was no longer visible on electrophoresisafter digestion with either DNAse or RNAse,suggesting that this contained both RNA andfragmented DNA.

PARAFFIN WAX EMBEDDED TISSUEDNAThe same procedures as above were inves-tigated with paraffin wax embedded tissuesections.Proteinase K: For paraffin wax embeddedmaterial, increasing the length of incubationwith the enzyme substantially improved theyield of DNA. A five day incubation producedalmost 10 times as much DNA as an overnightincubation (table 2), with a correspondingincrease in the molecular weight of the extrac-ted DNA (fig 2). A long proteinase K digestionwas also capable of extracting DNA from asingle section of paraffin wax embedded tissue.Dewaxing the tissue in xylene had no effect on

Table 1 DNA yieldfrom 100 mg blocks offresh tonsil, extracted using different optimised protocols

Proteinase K Proteinase K SDS incubation Boiling

(I hour) (18 hours) (I hour) (15 minutes)

Mean DNA yield 5 45 5 17 0-79 0-58(pg/mg)values (5 1,5-5,5-76) (4-22,5 62,5-66) (0 653,0 936,0 786) (0 616,0-211,0.924)

Mean yield as a percentage 100 94 9 14-5 10-6ofmost efficient method

n = 3: values in parentheses.



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


Table 2 DNA yieldfrom sectionst ofparaffin wax embedded tonsil, waxed or dewaxed in xylene,extracted by varying lengths ofproteinase K incubation

Mean DNA yield (pg/section)Length of incubation (37°C) (hours) Waxed Dewaxed

1 0-06 (0-02,0-03,0-14) 0 03 (0-01.0-03.0-04)3 0-04 (0 025,0 04,0 05) 0 07 (005,0 07,0 08)

24 0-22 (0-16,0-22,0-27) 0-27 (0-14,0-31,035)48 0-72 (0-44,0-77,0-95) 0-95 (0-51,1-14,1-21)120 2 09(1 29,2 30,2-67) 2-10 (1-49,2-30,2-52)

n = 3 values in parentheses.tTen serial sections from the same block of tissue.

the yield or quality of the DNA (table 2, fig 2).SDS: Increasing the digestion time from one to4 hours did not improve DNA yields.Boiling: Increasing the boiling time made noappreciable difference to the DNA yield.Comparison of paraffin embedded tissue DNAextraction techniques: A five day proteinase Kincubation produced greater yields than SDSincubation, but the simple rapid boilingmethod produced one fifth as much DNA asthe long enzyme incubation (table 3). TheDNA extracted by proteinase K was of a highmolecular weight, while the SDS and boilingmethods produced lower molecular weight andfragmented DNA. Some of theDNA extractedby boiling remained in the loading well of theelectrophoresis gel, suggesting that it may haveremained partially associated with protein (figlb).

RNAAfter digestion with either RNAse or DNAsethe small amount of low molecular weightmaterial extracted from paraffin wax embed-ded tissue by SDS, incubation, or boiling wasno longer visible on gel electrophoresis, againsuggesting that they contained both RNA andfragmented DNA. RNAse digestion of thesample produced by proteinase K incubation,however, showed that this procedure producedrelatively large amounts ofRNA.

DNA PCRThe DNA extracted from routinely processedspleen by all of the extraction methods was of aquality suitable for amplification by thepolymerase chain reaction. The DNA sampleswere of a sufficient concentration such that 1 iulof each could be used as template for amplifica-tion, producing a 142 base pair fragment of thefactor VIII gene (fig 3, lanes 2, 4, 5). Theextracted DNA sample from the spleen usingproteinase K was of a very high concentration(980 ng/pl), and from this and other results(not shown) it seems that adding more than500 ng of DNA as the PCR template mayactually inhibit the amplification reaction.Adding one tenth of the DNA (98 ng)

Table 3 DNA yieldfrom sectionsf ofparaffin wax embedded tissue extracted using differentoptimised protocols

Proteinase K SDS incubation Boiling(5 days) (I hour) (is mins)

Mean DNA yield (pg/section) (n = 2) 1-75 0-088 0-325Values 1-84,1-65 0-05,0 125 0-33,0 32Mean yield as a percentage of most

efficient method 100 5 18-6

tTen serial sections from the same block of tissue.

1 2..%.M.mm.

3 L 5 6 7

Figure 3 2% agarose electrophoresis gel ofPCRproducts, showing amplification of 142 base pairsfragment of FVIIIgenefrom DNA extractedfromparaffin wax embedded sections of spleen tissue bydifferent methods.(1) Standard: 123 base pair ladder, (2) proteinase Kincubation (five days); (3) 1/10 dilution ofproteinase Ksample; (4) SDS incubation (one hour); (5) boiling (15minutes); (6) intact, undigested, dewaxed section; (7)negative control: distilled water.

produced a much brighter more distinct PCRproduct band (fig 3, lane 3), compared with thefainter band and non-specific smear whenusing larger amounts of template (fig 3, lane 2).Amplification of the factor VIII gene was alsoattempted using an intact, non-digested,dewaxed section of paraffin wax embeddedspleen as template. This reaction did notproduce the factor VIII specific 120 base pairfragment (fig 3, lane 6).

FIXATIONBlocks of fresh tonsil were left in sealed con-tainers at room temperature for up to sevendays before fixation for 24 hours in 4%O formal-dehyde, and the DNA extracted using anovernight proteinase K incubation.Surprisingly, the yields obtained did notdecrease with an increasing delay before fixa-tion, suggesting that leaving tissue unfixed forup to seven days caused only a negligibledegradation of DNA, compared with the fixa-tion process itself. Fresh tissue producedalmost 40 times as much DNA (5.014 jg/mg)as tissue fixed immediately in 4%O formalde-hyde for 24 hours (0-135 jg/mg), and theDNAwas of very high molecular weight, comparedwith the fragmented material extracted fromfixed tissue.

Leaving tissue in fixation for an extendedperiod (up to 12 weeks), resulted in a furtherdecrease in the amount ofDNA that could beextracted (fig 4). After an initial 24 hours'fixation at 37'C, however, the yields of RNAextracted after two to three weeks in fixativewere much higher than if the fixative was keptat room temperature, while stronger heating(up to 60'C) during the first 24 hours resultedin much poorer yields ofDNA. After 12 weeks

502



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Tissue extraction ofDNA and RNA and analysis by PLR

Figure 4 Yield ofDNAextractedfrom 100 mg oftonsilfixedfor varyingperiods informalin afteran initial 24 hours'fixation at roomtemperature, 37C or60°C.

0.20-0.18--16--0-1L

z 010; 0.08-:X 0.06-*s 0*04

0.02I_

1 2 3 4 5 6 7 8Initial temperature of fixation

Room temperature

370C=---" 600C

0 2 A 6 8Duration in fixature (weeks)

10 12

in fixative the yields ofDNA extracted from thetissue were low and independent of the initialtemperature of fixation.The yields of DNA extracted from tissue

fixed in 4% formaldehyde and a variety ofother commonly used fixatives were compared.Tissue fixed in formaldehyde gave more DNAthan tissue fixed in other formaldehyde basedfixatives, paraformaldehyde and Bouin'sreagent (table 4). The DNA extracted fromthese tissues was of very low molecular weight(fig 5, lanes 4-8). Surprisingly, it was possibleto extract some DNA from material fixed inBouin's reagent, a picric acid based fixative.Tissue fixed in Carnoy's reagent producedyields of DNA 20 times those from formalde-hyde fixed tissue and this DNA was of a muchbetter quality, containing some high molecularweight fragments (fig 5, lanes 2 and 3). DNAPCR of the extracted DNA was successful onfive of seven samples, including that fixed inBouin's reagent. DNA from the tissue fixed informol sublimate or paraformaldehyde couldnot be amplified. This may have been due to aninhibition of the Taq polymerase enzyme byone of the constituents of these fixatives-forexample, by the mercury contained in formolsublimate.

DiscussionThe use of a quantitative DNA assay haspermitted the critical analysis of a variety ofDNA extraction techniques and their applica-tion to different tissue preparations.For the extraction ofDNA from fresh tissue,

we have found that routinely quoted protocolsmay not be the most efficient as a one hourproteinase K incubation produces as muchDNA as longer incubations. This only partlyreduces the time taken for extractionprocedures in fresh tissue, however, as theDNA produced by this method is of highmolecular weight and may take up to three daysto solubilise fully. Of other reported methods,the one hour SDS incubation produces pooreryields of DNA and offers no advantages over

Table 4 DNA yieldfrom 100 mg blocks of tonsilfixed in a range of differentfixative agents

Mean DNA yieldDuration of Mean DNA yield as a percentage of

Fixative fixation (hours) (pg/mg) optimal method

Carnoy's 5 2-575 (3-29,1-86) 100Carnoy's 24 1-805 (1-89,1-72) 70Tap water formalin 24 0-089 (0 11,0 068) 35Neutral buffered formalin 24 0-061 (0-059,0-063) 24Formol sublimate 24 0-068 (0-083,0-053) 26Paraformaldehyde 24 0-063 (0-053,0-072) 24Bouin's fixative 24 0-046 (0-053,0-039) 18

n = 2: values in parentheses.

Figure 5 0.7% agarose electrophoresis gel ofDNAextractedfrom tissuefixed in differentfixative agents byovernight proteinase K incubation. (1) Standard: ADNA Hind III digest; (2) Carnoy's reagent (fivehours); (3) Carnoy's reagent (24 hours); (4) tap waterformalin; (5) neutral bufferedformalin; (6) formolsublimate; (7) paraformaldehyde; (8) Bouin's reagent.

enzyme digestion. While simply boiling tissueis a very rapid procedure that requires nofurther purification of the DNA, the DNA isnot fully dissociated from tissue proteins andthe yields of extracted material are low, of poorquality, and unlikely to be suitable for manymolecular techniques, with the exception of thepolymerase chain reaction.The extraction of genetic material from

paraffin wax embedded tissue is very importantas it permits the retrospective molecular inves-tigation of the aetiology and epidemiology ofdisease. Our study on the effect of fixationprocedures on DNA extraction has quan-titatively shown that it is the fixation of tissue,rather than any delay in fixation or furtherprocessing, which prevents the extraction ofhigh molecular weight DNA using a shortproteinase K incubation, as suggestedpreviously.'2 The yield and quality of DNAfrom formaldehyde fixed tissue, however, canbe improved by increasing the proteinase Kdigestion to five days. The DNA extracted bythis method is likely to be suitable for somemolecular techniques and certainly for theamplification of a single copy gene using PCR.DNA samples extracted by this method mayrequire dilution before use with PCR as theamplification may be inhibited by too muchtemplate DNA. Alternatively, a product of thefixation or processing procedure may be co-extracted along with the DNA, and this may beresponsible for the inhibition of amplification,because this effect is not seen with DNAextracted from fresh tissue (unpublishedobservations). A long proteinase K digestion isan efficient extraction procedure and is capableof extracting DNA from as little as one paraffinwax embedded section, and so can be used toproduceDNA from small quantities ofmaterialfor PCR analysis.

-I-.l

503



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


The effects of fixation on DNA extractioncan be substantially reduced by using a fixativeother than formaldehyde. Carnoy's reagentresults in a much better preservation ofDNA inthe tissue, allowing the extraction of highmolecular weight DNA"3 and quantitativelymuch higher yields, using more rapidextraction procedures. Although Carnoy's is apoor fixative for histological purposes, itmay be potentially useful for the fixation oftissue collected specifically for molecularinvestigations where low temperature storage isnot available.Although an efficient and sensitive extrac-

tion technique, a long proteinase K incubationmay take at least seven days to produce a usableDNA solution. Boiling sections for just 15minutes produces relatively good yields fromparaffin wax embedded tissue and a DNAsolution of sufficient concentration to use a 1 4ulaliquot as template for the amplification of thesingle copy factor VIII gene. This method,however, is unlikely to produce enough DNAfor the investigation of low copy number gen-etic material as is found in some persistent viralinfections. This could be achieved by increas-ing the sensitivity of the amplification reactionby increasing the number of cycles, althoughthis has the disadvantage of reducing thespecificity of the reaction with the productionof extraneous bands. Alternatively, a moresensitive autoradiographic detection methodcould be used, but this would reduce thesimplicity and rapidity of the technique. Theuse of an intact paraffin wax embedded sectiondid not permit amplification of factor VIII, asingle copy gene, although it has been used toamplify fragments of the Alu repeat sequence(unpublished results). This suggests that a highcopy number of target sequences must bepresent in the tissue if it is to be used directly asa template for amplification, with no extractionof nucleic acid, using less than 40 cycles ofPCR.RNA extraction procedures tend to be com-

plex to avoid RNAse contamination and havenot been described for use with paraffin waxembedded material. We have shown that a

relatively simple and convenient technique,involving digestion with proteinase K, canextract considerable amounts of RNA fromfixed tissue. RNA extracted by this method issuitable for the PCR analysis of viral sequen-ces." This allows one simple procedure to beused to extract both DNA and RNA from atissue sample, making more tissue available formolecular investigation.

We thank the Jean Shanks Foundation for supporting D PJackson. This work was partly funded by grants from theYorkshire Cancer Research Campaign and the Crohn's Diseasein Childhood Research Appeal.We are indebted to Miss J I Hamblin and Mrs J Fearnley for

their help in preparing the manuscript.

I Strauss WM. Preparation of genomic DNA from mam-malian tissue. In: Ausubel FK, Brent R, Kingston RE, etal, eds. Current protocols in molecular biology, 1987-8. NewYork: Greene Publishing Associates& Wiley Interscience,1987:2 2 1-2 2 3.

2 Maniatis T, Fritsch EF, Sambrook J. Molecular cloning: Alaboratory handbook. Cold Spring Harbor Laboratory,New York: Cold Spring Harbor, 1982.

3 Goelz SE, Hamilton SR, Vogelstein B. Purification ofDNAfrom formaldehyde fixed and paraffin embedded humantissue. Biochem Biophys Res Commun 1985;130:1 18-26.

4 Dubeau L, Chandler LA, Gralow JR, Nichols PW, JonesPA. Southern blot analysis of DNA extracted fromformalin-fixed pathology specimens. Cancer Res1986;46:2964-9.

5 Chirgwin JM, Przybyla AE, MacDonald RJ, Rutter WJ.Isolation of biologically active ribonucleic acid fromsources enriched in ribonuclease. Biochem 1979;18:5294-9.

6 Saiki RK, Scharf S, Faloona F, et al. Enzymatic amplifica-tion of fl-Globin genomic sequences and restriction siteanalysis for diagnosis of sickle cell anaemia. Science1 985;230: 1350-4.

7 Hart C, Spira T, Moore J, et al. Direct detection of HIVRNA expression in seropositive subjects. Lancet1 988;ii:596-9.

8 Lynas C, Cook SD, Laycock KA, Bradfield JWB, MaitlandNJ. Detection of latent virus mRNA in tissues using thepolymerase chain reaction. J Pathol 1989;157:285-9.

9 Hopwood D. Fixation and fixatives. In: Bancroft JD,Stevens A, eds. Theory and practice of histological tech-niques. 2nd ed. Edinburgh: Churchill Livingstone,1982:20-40.

10 Lench N, Stanier P, Williamson R. Simple non-invasivemethod to obtain DNA for gene analysis. Lancet1988;i: 1356-8.

11 Kogan SC, Doherty M, Gitschier J. An improved methodfor prenatal diagnosis of genetic diseases by analysis ofamplified DNA sequences. N Engl J Med 1987;317:985-90.

12 Burmer GC, Rabinovitch PS, Loeb LA. Analyis of c-Ki-rasmutations in human colon carcinoma by cell sorting,polymerase chain reaction and DNA sequencing. CancerRes 1989;49:2141-6.

13 Jackson DP, Quirke P, Lewis F, et al. Detection of measlesvirus RNA in paraffin-embedded tissue. Lancet1989;i: 1 391.

504



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

laboratory techniques tissueextractionof · stat frozen sections (8 pm) or paraffin wax...

Documents